LATE BRONZE AGE MARITIME TRADE IN THE EASTERN MEDITERRANEAN :

Uppsala University Department of Archaeology and Ancient History Uppsala University LATE BRONZE AGE MARITIME TRADE IN THE EASTERN MEDITERRANEAN : AN INLAND LEVANTINE PERSPECTIVE Kristina Josephson Hesse Classical Archaeology and Ancient History Master thesis. 2008

ii ABSTRACT Josephson Hesse, K., 2008. Late Bronze Age Maritime Trade in Eastern Mediterranean: An Inland Levantine Perspective. Master thesis. Department of Archaeology and Ancient History. Uppsala University. This paper emphasizes the nature of trade relations in the Eastern Mediterranean in general and from a Levantine inland perspective in particular. The maritime trade relation of the ancient city of Hazor, located in the interior of LB Canaan is a case study investigating the Mycenaean and Cypriot pottery on the site. The influx of these vessels peaked during LB IIA. The distribution and types of this pottery at Hazor point to four interested groups that wanted it. These were the royal and religious elites; the people in Area F; the religious functionaries of the Lower City; and the craftsmen of Area C. The abundance of imports in Area F, among other evidence, indicates that this area might have contained a trading quarter from where the imports were distributed to other interested groups. A model of interregional interaction networks, which is a modified world systems approach, is used to describe the organization of trade connections between the Levant, Cyprus and the Aegean and even beyond. The contents of the Ulu Burun and Cape Gelidonya ships, wrecked on the coast of south Turkey, show that luxury items were traded from afar through Canaan via the coastal cities overseas to the Aegean. Such long-distance trade with luxury goods requires professional traders familiar with the risks and security measures along the routes and with the knowledge of value systems and languages of diverse societies. These traders established networks along main trade routes and settled in trading quarters in particular node cities. The paper suggests that Hazor, as one of the largest cities in Canaan, located along the main trade routes, possessed such a node position. In this trade the Levantine coastal cities of Sarepta, Abu Hawam, Akko and possibly Tel Nami seem to have played important roles. These main ports of southern Syria and northern Palestine were all accessible to Hazor, although some of them in different periods of LB. Keywords: Hazor, Mycenaean pottery, Cypriot pottery, Bichrome pottery, maritime trade, Eastern Mediterranean, Late Bronze Age, cultural interactions, -relations, - contacts, - exchange, Syria-Palestine, Canaan, the Levant Kristina Josephson Hesse, Department of Archaeology and Ancient History, Uppsala University. Box 626, 75126 Uppsala. Sweden.

TABLE OF CONTENTS ABSTRACT... II 1 INTRODUCTION... 2 1.1 BACKGROUND - HAZOR S POSITION IN A LATE BRONZE AGE CONTEXT... 2 1.2 AIM AND QUESTIONS... 5 1.3 METHODS AND THEORIES... 5 1.3.1 Criticism of the Sources... 7 1.4 PREVIOUS RESEARCH... 8 2 THE INFLUX AND USE OF MYCENAEAN AND CYPRIOT POTTERY IN THE LEVANT DURING THE LATE BRONZE AGE... 10 2.1 CHRONOLOGY IN THE LEVANT, CYPRUS, AEGEAN AND EGYPT... 10 2.2 CYPRIOT AND MYCENAEAN POTTERY IN THE LATE BRONZE AGE LEVANT... 11 2.2.1 Cypriot Pottery... 11 Cypriot types and forms... 11 Influx and distribution of the Cypriot pottery... 13 2.2.2 Mycenaean Pottery... 14 Mycenaean types and forms... 14 Influx and distribution of Mycenaean pottery... 16 2.2.3 Provenance Studies and Imitations... 17 The Origin of the Bichrome Ware... 19 3 THE DISTRIBUTION OF IMPORTED POTTERY AT HAZOR... 20 3.1 AREA DESCRIPTIONS... 20 3.2 THE LB I CYPRIOT POTTERY... 22 3.2.1 The Bichrome Pottery at Hazor... 23 3.3 THE LB II CYPRIOT AND MYCENAEAN POTTERY... 24 3.3.1 Some Notes on the Imported Pottery of the Renewed Excavations... 27 Area A The Ceremonial Palace... 27 Area M The Podium Compound... 28 4 SHAPE, CONTENT, CONTEXT AND USE OF THE IMPORTED POTTERY AT HAZOR... 30 4.1 WHITE SLIP BOWLS AND BASE RING JUGS MOST COMMON CYPRIOT SHAPES AND TYPES... 30 4.1.1 Importers and Users of the Cypriot Pottery at Hazor... 32 4.2 FINE TABLE WARE AND TRANSPORT CONTAINERS FROM THE AEGEAN... 32 4.2.1 The Users of the Mycenaean Pottery at Hazor... 35 5 LEVANTINE HARBOR CITIES ACCESSIBILITY AND IMPORTS... 38

iv 5.1 THE LEVANTINE TERRESTRIAL COMMUNICATION NETWORK... 38 5.1 NORTHERN LEVANTINE HARBOUR SITES... 40 5.1.1 Ugarit / Minet el Beida... 40 5.1.2 Tell Kazel (ancient Sumur)... 42 5.1.3 Sarepta... 44 5.1.4 Some notes on other Northern Levantine coastal sites... 45 5.2 SOUTHERN LEVANTINE HARBOUR SITES... 45 5.2.1 Tell Abu Hawam... 45 5.2.1 Akko and Tel Nami... 47 5.2.2 Other Harbor Cities along the Southern Levantine Coast... 49 5.3 CONCLUSION LEVANTINE HARBOUR CITIES... 49 6 CIRCULATION, TRADE AND EXCHANGE... 51 6.1 THE ORGANIZATION OF THE EASTERN MEDITERRANEAN TRADE... 51 6.1.1 Substantivists vs. Formalists... 51 6.1.2 Complex Network Systems of Multifaceted Economics... 52 6.1.3 The International Character of Trade and Exchange in Eastern Mediterranean... 53 Sea routes... 54 6.2 LEVANTINE MARITIME TRADE RELATIONS... 56 6.2.1 Exchange products... 58 6.2.2 Trading quarters/diasporas... 59 6.3 THE IMPORTANCE OF TRADE AT HAZOR... 60 7 SUMMARY AND CONCLUSION - THE NATURE OF HAZOR S MARITIME TRADE RELATIONS... 62 APPENDIX... 66 TABLE I MYCENAEAN POTTERY AT HAZOR... 66 TABLE II CYPRIOT POTTERY AT HAZOR... 69 BIBLIOGRAPHY... 74 LIST OF ILLUSTRATIONS... 87

1 ABBREVIATIONS Following abbreviations are used in the text: BR Base Ring EBR Early Base Ring FM Furumark Motifs FS Furumark Shape LBA Late Bronze Age LC Late Cypriot LH Late Helladic LM Late Minoan MBA Middle Bronze Age MM Middle Minoan NAA Neutron Activation Analysis PBR Proto Base Ring PPI Peer Polity Interaction PWS Proto White Slip WS White Slip WST World Systems Theory

2 1 INTRODUCTION 1.1 BACKGROUND - HAZOR S POSITION IN A LATE BRONZE AGE CONTEXT The Canaanite city of Hazor was located about 15 km north of the Sea of Galilee. The designations of this region, used by archaeologists, sometimes cause some confusion because of their present day geographical, cultural or political connotations. However, commonly accepted terms along with Canaan are the Levant or Syria-Palestine, where Syria characterizes the ancient northern and Palestine the ancient southern part of this geo-cultural area. The city-state Hazor was located between the northern and southern part of the region (Fig. 1). 1 Fig. 1. Hazor in Canaan 1 The Levant includes parts of present-day Turkey, Syria, Lebanon, Israel, Jordan and the Palestine Autonomy (Wijngaarden van 2002, 31). LB Canaan was somewhat smaller and extended from Wadi al- Arish (the brook of Egypt, just south of Gaza) to the vicinity of Ugarit in northern Syria. It is mentioned in Egyptian sources, several Akkadian documents, statue inscriptions etc (Hacket 1997, 409 and references therein; Rainey 2003, 169-172). LB Canaan should be seen as a city-state culture, comprising several interacting city-states, among them Hazor, in the definition of C. Renfrew s Early State Module (Renfrew 1986, 2; Zuckerman 2003b, 216 in Hebrew).

3 Excavations of Tel Hazor were conducted by Y. Yadin in 1955-1958 and 1968, 2 followed by renewed excavations by A. Ben-Tor from 1990 onwards. 3 The city consisted of two distinct parts; the Upper City, which is the tell proper, and the Lower City comprising a vast plateau to the north and partially to the east of the tel (Fig. 2). Fig. 2. Tel Hazor with Area designations, aerial photo (looking north) The Upper City, which was occupied from the Early Bronze Age to the Persian period, had impressive fortifications and public buildings and was densely populated. The Lower City was well fortified by natural slopes on the east, a deep and narrow valley on the 2 Reports in Ben-Tor & Bonfil 1997; Yadin, et al. 1989; Yadin, et al. 1958; 1960; Yadin, et al. 1961. 3 No final reports from the renewed excavations are yet published, however preliminary results and a couple of extended summations are available see Ben-Tor 1992-2007; 1999; Ben-Tor & Rubiato 1999. The so far most extensive research concerning the LB levels of the renewed excavations is made by S. Zuckerman in her unpublished Ph.D. thesis (2003a; 2003b in Hebrew).

4 north and a manmade moat and earthen rampart on the southwest side of the plateau. Excavations in this area yielded remains of temples, dwellings and well-planned sewage systems. 4 The occupation of the Lower City is dated from approximately the 18 th to the 13 th century B.C. (MB II LB II). This period constituted the heyday of the city. 5 During the Late Bronze Age the city-states of Canaan were dominated by Egypt. 6 Across this land-bridge, main roads, fortifications, granaries and water installations supplied the Egyptian troops and facilitated their movements to their enemies in the north, the Hittites and the Mitannians. 7 In addition the pharaonic court levied tribute on the Canaanite city-states and requested different kinds of goods to be sent to Egypt. Egyptian lists of tribute and booty, especially those of Thutmosis III (1479-1425 BC), as well as Akkadian documents, primarily the Amarna letters, provide information about these goods. 8 Hazor s important position on the crossroad of the main inland trade routes between Egypt and Mesopotamia and between the coast and the Transjordanian desert promoted contacts and exchange with different cultural spheres, which also possessed different realms of raw material. This paper will mainly focus on aspects of the imported Cypriot and Mycenaean pottery found on the site. Cypriot pottery has been uncovered in MB IIB contexts at Hazor, as well as a few sherds of Kamares Ware from Crete, dated to MM IIB or IIIA (the end of MB II). It is suggested that these types indicate indirect contacts with those islands as part of the relations with Levantine coastal cities such as Ugarit and Byblos. 9 The imported pottery of interest in this paper derives, however from the Late Bronze Age, generally seen as an era of internationalization because of the prevailing flourishing trade and intercultural contacts throughout the Eastern Mediterranean. 10 4 Ben-Tor & Bonfil 1997; Yadin 1972; 1975; Yadin, et al. 1989; Yadin, et al. 1958; 1960. 5 Ben-Tor & Bonfil 1997, 13. 6 The degree of this domination fluctuated through times. The scholarly opinions differ regarding the nature of this domination, see e.g the views of Higginbotham 2000; Mazar 1992; Weinstein 1981. 7 Cohen 2000, 87; Helck 1962, chap. 27; Na'aman 1981, 177-181; 2000, 131f; Weinstein 1981. 8 Ahituv 1978; Moran 1992, xxvii; Na'aman 1981, 173f. 9 Dothan, et al. 2000, 13. 10 See e.g Nordquist 2007.

5 1.2 AIM AND QUESTIONS Mycenaean and Cypriot pottery constitute a category of externally influenced material found at Hazor. An investigation of these pottery types and their implications will be offered in this paper as well as an investigation of trade routes and Levantine harbour cities, aiming to disclose the nature of Hazor s (indirect) maritime trade connections as well as some aspects of the Mycenaean and Cypriot trade in the Eastern Mediterranean in general. Questions of importance are related to the influx of pottery types and forms in the Levant and particularly at Hazor. Why was the pottery imported to Hazor? What was its find context in the city? Who used this pottery and how? Other questions of importance are associated with the infrastructure, for example the location of the city in relation to trade routes and harbours. 1.3 METHODS AND THEORIES The special type of view adopted in this paper, viz an inland site s connection to maritime trade, offers an extended understanding of the Late Bronze Age trade between the Aegean, Cyprus and the Levant, since it explains aspects (social, economic and cultural) of the movement of goods not only overseas but also from the maritime harbours to, and beyond, the Levantine interior and vice versa. The interior, in this paper, is exemplified by the city of Hazor and the movement of goods is studied through an investigation of accessible trade routes and harbours. The distribution and use of the imported Mycenaean and Cypriot pottery on this site will hence be considered as well as possible exchange goods the other way around. P. Nick Kardulias discussion on multiple levels in the Aegean Bronze Age World System is, in part, applicable to the particular inland Levantine perspective of this paper. He observes three levels of interaction spheres in the Aegean: 11 1) the internal system which involved the exchange of goods on a local level, e.g. between the small states within Crete or Argolid; 2) an intermediate system of interaction between the Aegean islands and/or between the islands and the mainland; 12 3) a system of long-distance 11 Kardulias 1999, 187-192. 12 Both these spheres included entities on rather equal stature, which he compares to the small states within an Early State Module according to Renfrew s description (1975; 1986.)

6 interactions which included connections with the Near East, Anatolia and Egypt. Kardulias explains the Eastern Mediterranean World System as: an international interchange that involved the transfer of both bulk goods and preciosities to and from the Aegean, Egypt, the Syro-Palestinian coast, Cyprus, and Anatolia. Direct Aegean contacts were limited to coastal areas in these other lands. The network, however, did extend far beyond the littoral zones. 13 By applying Kardulias concept, the interaction sphere of Hazor can be divided into three levels as well: a) a local system which included the exchange of goods (possibly redistributive) between the city of Hazor and the subordinated towns and villages in its fertile hinterland; 14 b) a regional system which operated within the Canaanite city-state culture, comparable to C. Renfrew s Early State Module; 15 c) an interregional system that places Hazor within an Eastern Mediterranean World System of trade and interaction. It is mainly this last level, although in a modified way, that will be described in this paper since it focuses on the inter-regional trade of Hazor as part of a wider long-distance distribution network in the Eastern Mediterranean and even beyond. The original model of the World Systems Theory (WST) was developed for modern societies and was based on the current political and economic structure of capitalist world domination. 16 This domination is maintained by economic processes that link regions together in dependent relationships through the exploitation of raw-material rich peripheries by dominating cores. 17 Since WST is a modern concept, it has been up to the archaeological establishment, together with sociologists, to adjust it to the ancient world and archaeological demands. 18 A matter of adaptation has been Wallerstein s assumption that pre-capitalist societies were not part of economic systems and thus not suitable for WST. This is however not a problem if adopting a formalist view of ancient 13 Kardulias 1999, 195. 14 The nature of this local exchange is hard to investigate since an administrative LB palace is not yet unearthed at Hazor. 15 Renfrew 1986, 2; Zuckerman 2003b, 216 (in Hebrew). 16 Wallerstein 1974. 17 Wallerstein 1974, 349f. 18 See e.g. Champion 1989; Chase-Dunn & Hall 1993; Edens 1992; Kohl 1989; McGuire 1986; 1989; Rowlands, et al. 1989; Stein 1999b.

routes. 22 The kind of interactions described in this paper are mainly core-to-core based, paper. 23 It should also be noted that the number of sites excavated, and thus pottery found, 7 economics. P. Kohl has convincingly demonstrated the market mentality and complexity of ancient economics in his revision of WST. 19 Another matter of adaptation has been the erroneous supposition that core dominance denies any kind of agency to the periphery, an assumption derived from the economic determinism of the World Systems model and its sole focus on the core activities. 20 This, together with other drawbacks of WST, caused G. Stein to develop a theory, applicable to ancient economics, based on the more flexible term interregional interaction networks, which were operated by trade groups/diasporas that established networks along major trade routes. 21 Inspired by P. Curtin and A. Cohen he describes these systems as operated by trade groups/diasporas that established along major trade since the relations between the Levant, Cyprus and the Greek mainland should be seen as operating on a rather equal basis. In short, an interregional network model will be used in this paper to explain Hazor s indirect long-distance trade relations via the Levantine coastal cities. This trade should be regarded as part of a modified perspective on an Eastern Mediterranean World System. 1.3.1 Criticism of the Sources The pottery and finds of the renewed excavations at Tel Hazor (Area A and M) are to a certain extent considered in the discussions below; however, since the examination of these findings is not yet finished and is only briefly described in preliminary reports and articles, it is not possible to include it in the catalogues and statistical diagrams of this in the southern Levant is much higher than in the northern part, which is caused by 19 Kohl 1989. Regarding formalists and substantivists view on ancient economy see Chapter 6. 20 Stein 1999a, 42f; 1999b, 155. 21 Stein 1999a, 46-48. 22 Curtin 1984.Cohen, A 197? 23 The Hazor team is currently producing the final reports, which will be published within a couple of years.

8 scientific traditions as well as historical efforts invested in exploring the biblical sites of the former. 24 In addition the political situation has prevented sufficient excavations of the important coastal cities of modern day Lebanon. 25 Such factors are considered as much as possible in this paper when comparing the amount of pottery and other aspects of the cities in different parts of the Levant. 1.4 PREVIOUS RESEARCH The increasing research field dealing with trade relations in the Eastern Mediterranean has resulted in several conferences and studies these last decades. 26 This attention to ancient maritime trade involving the Aegean, Cyprus and the Levant has, to a great deal, been sponsored by the discoveries of the Cape Gelidonya and the Ulu Burun shipwrecks. 27 Pottery constitutes the cultural and chronological framework of trade and interaction studies because of the possibilities of determining its origin, date and use as well as its advantages of being a lasting material. In this sense the fine Mycenaean and Cypriot wares imported to the Levant are highly essential. The primary study of the Mycenaean pottery was made by A. Furumark. He divided the then known material into 336 vessel-shapes (FS= Furumark Shape) and the painted decorations into 78 motifs (FM=Furumark Motif), which all were assigned a place within a framework of relative chronology. 28 In the 1920 s, E. Gjerstad already regarded the relations between Cyprus and Palestine as commercial and all the Cypriot pottery in the southern Levant as imported. In addition he suggested that commercial agencies existed along the Palestinian coast who were supplied by Cypriot ships charged with wares which were then transported inland by caravan, 29 E. Sjöquist observed in 1940 s the distribution of LC I pottery in southern- and LC II pottery in northern and central Palestine with Megiddo and Beth-Shean as transit centers. 30 Today, however, 24 Hankey 1993, 101; Wijngaarden van 2002, 16. 25 Seeden 1990, 5f. 26 E.g. Cline 1994; Cline & Harris-Cline 1998; Gale 1991; Laffineur & Greco 2005. 27 Bass 1967; 1986; 1998; 2005; Pulak 2005a; 2005b. 28 Furumark 1941a; 1941b. 29 Gjerstad 1926, 310-312. 30 Sjöqvist 1940.

9 additional sites have been excavated and probable emporia, for both Mycenaean and Cypriot pottery, on the Levantine coast are identified (see Chapter 5 below). Sjöquist also created long and useful lists of all the Cypriot pottery found in Palestine. These lists were later supplemented by P. Åström in 1972, 31 and in a Ph.D study by B. Gittlen in 1977, who contributed a great deal to the understanding of the distribution of this pottery in Palestine. 32 Also, the importance of V. Hankey s studies on Mycenaean pottery and trade in the Levant cannot be overestimated. 33 Her work on Mycenaean finds in the Middle East were updated and extended in 1994 by A. J. Leonard. 34 These chronological frameworks and lists function as a base for more developed and recent studies regarding different aspects (e.g. reasons of importation, main influx, trade and possible emporia) of the interactions between the Levant and the Aegean and Cyprus. 35 A recent tendency towards a more contextual archaeology concerning the interpretation of imported pottery within certain sites in the Levant can be observed. 36 The Mycenaean pottery at Hazor has earlier been given scholarly attention. In G. J. van Wijngaarden s book on Mycenaean pottery in the Levant, a contextual investigation of the distribution and use of Mycenaean pottery at Hazor was conducted, which has been useful for this study; however, his investigation did not deal with the city s trade relations or how and why the pottery reached Hazor, neither did he discuss the Cypriot pottery. 37 A subchapter of S. Zuckerman s PhD thesis is dedicated to the occurrence of Mycenaean and Cypriot pottery at the site. Her focus is however solely on the chronological aspects of the pottery in discussions related to the dating of the destruction stratum of the final Late Bronze Age phase. 38 31 Åström 1972. 32 Gittlen 1977 (1984). 33 Hankey 1967; 1981; 1993. 34 Leonard 1994. 35 See e.g. Gittlen 1981; Hankey 1993; Killebrew 1998; Leonard 1981; Prag 1985; Sherratt 1999; Wijngaarden van 2002. 36 Steel 2002; Wijngaarden van 2002. 37 Wijngaarden van 2002, 75-79. 38 Zuckerman 2003b, 269-281 (in Hebrew).

10 2 THE INFLUX AND USE OF MYCENAEAN AND CYPRIOT POTTERY IN THE LEVANT DURING THE LATE BRONZE AGE 2.1 CHRONOLOGY IN THE LEVANT, CYPRUS, AEGEAN AND EGYPT Fig. 3. Comparative chronology chart of Eastern Mediterranean areas The chronological division in the table above (Fig. 3) should be used as a guideline in this paper. It is based on suggestions made by scholars dealing with chronologies from different areas in the Eastern Mediterranean. The Aegean and Cypriot chronology adopted in the chart derive from Warren and Hankey. 39 Their discussion regarding absolute dates refers to the Egyptian chronology made by Kitchen. Hence, the dates of 39 Warren & Hankey 1989, 118, 138, 169.

11 pharaohs mentioned in this text follow his suggestions. 40 The Levantine LBA is dependent on the Egyptian chronology as well. LB I starts with the beginning of the 18 th Dynasty and the expulsion of the Hyksos from Avaris by Ahmose I. LB IIA is the second half of the 18 th Dynasty, mainly the Amarna period. LB IIB correspond to the 19 th Dynasty and ends with the Egyptian decline and withdrawal from Canaan and the invasion/immigration of new groups of people especially along the coastline of northern and southern Palestine. 41 2.2 CYPRIOT AND MYCENAEAN POTTERY IN THE LATE BRONZE AGE LEVANT 2.2.1 Cypriot Pottery Cypriot types and forms The Cypriot ware found in the Levant (Fig. 4) include mainly the Monochrome bowls, Bucchero jugs, White Shaved jugs and juglets, White Slip bowls (the so called milk bowl ), and zoomorphic figurines, bowls, jugs and juglets of Base Ring ware. Of these BR II (especially the jugs called bilbils ) and WS II ( milk bowls ) were the most widely distributed types. 42 WS Bowls were most common in habitation context while BR juglets and jugs, which constitute the most frequent form, occurred mainly in funerals. 43 This contrasts to Cyprus where bowls are the most common form in LC burials. The export of a large amount of jugs and juglets and the occurrence of less decorated and less elaborated LC forms in Palestine compared to Cyprus, in combination with the invention of White Shaved juglets for Palestinian consumption made Gittlen propose a selective trade which was based upon the thorough knowledge of the Palestinian market. 44 40 Kitchen 1987; 1989; 2000, 49. 41 Gonen 1992. 42 The investigation dealing with the Cypriot pottery is mainly based on Gittlen s research on sites in Palestine (1977 (1984); 1981). Although no thorough investigation and compilation on the Syrian LC material has been conducted most of his results could be adopted on the northern part of the Levant as well, see e.g. Yon 2001. 43 Gittlen 1977 (1984), 510f; 1981, 52. 44 Gittlen 1977 (1984), 513; 1981, 52f.

Fig. 4. Cypriot pottery 12

13 Bichrome ware is usually described apart from the ordinary LC assemblage of the Levant because of the problematic issue of its origin and influence and how to separate the imports from the locally produced vessels of this type (see below Chapter 2.2.3). This ware includes mainly storage jars, jugs and kraters which are decorated in black and red. It is suggested that the distinct colors and the appearance of the decoration on the upper part and neck signifies a Cypriot production. 45. Influx and distribution of the Cypriot pottery The occurrence of Bichrome in Palestine is limited between late MB and early LB; therefore these vessels are important tools for dating. 46 Monochrome is another type appearing in MB II. BR, Bucchero, WS and White Shaved arrived during LB I. The earlier types of these (Proto BR, Monochrome, Proto WS and WS I) ceased before the entrance of LB II while BR I, BR II, Bucchero, White Shaved and WS II reached their zenith in LB IIA, the Amarna period. The trade during this period was completely dominated by BR II pottery (46 %). 47 The Cypriot pottery is distributed along the coastal cities as well as in the interior of Palestine. Some types (PBR, Early BR II, Bucchero, PWS) appear even more frequently on inland sites, which might depend on chronological reasons. Abu Hawam was, for example, not an important trade center until LB II. The distribution map of the Bucchero ware (LB II) shows a very odd picture however; it appears at six inland sites but only one coastal site, Tell el Ajjul in southern Palestine. 48 According to an investigation cited by Åström the distribution of WS II reveals a change in trade pattern within the Levant. The ties between Cyprus and northern Levant were stronger in the first and last phase of WS II, while the Cypro-Palestinian trade appears to have been more intense in the middle phase. 49 According to a brief investigation by Yon, the large scale importation of Cypriot pottery in the northern Levant was concentrated in the coastal sites. She claims that the 45 For the difference between imported and local Bichrome ware see Artzy, et al. 1978, 103. However their suggestions are not unequivocal (see Chapter 2.2.3). 46 Prag 1985, 161. 47 Gittlen 1981, 50f. 48 See Gittlen 1977 (1984), map 1-14. 49 Åström 1993, 309.

14 Cypriot pottery at e.g. Ras Shamra was mainly destined for local consumption or for distribution amongst the coastal population who were accustomed to this category of product. Inland redistribution did occur, although it was minimal. 50 However, it is hard to accept that the distribution of pottery differs so hugely between the inland of the northern and southern Levant even considering the separating mountain chain running along the Lebanese and Syrian coast, which actually has gaps, passes and valleys located where the largest coastal cities were located (see chapter 5 below). More likely, the differences depended on the less explored and published sites of the northern Levant, especially in its interior. The scarcity of Cypriot imports in LB IIB Levant caused Gittlen to suggest that this trade ceased in the beginning of LB IIB. 51 This fact is however questioned by some other scholars, who suggest that the trade continued throughout the Cypriot production period of the vessel types. 52 However, as stated by Gittlen, the Cypriot ware might have been used long after its importation ceased and caution should be taken when using this pottery for dating. 53 2.2.2 Mycenaean Pottery Mycenaean types and forms The most common Mycenaean vessels found in Syria-Palestine (Fig. 5) belong to the group of closed forms, which include stirrup jars, flasks, pithoid jars, square-sided alabastron, piriform jars and pyxis. About 60% of the imports belong to this group and hence 40% constitute the open forms. Of the open forms small vessels like chalices, cups and mugs were common, as well as large vessels like deep bowls and kraters. 54 Except for the classification of open or closed the Mycenaean vessels in the Levant can also be categorized into three main groups depending on their usage: Storage vessels (jars, small wide-mouthed containers, stirrup jars and flasks); dinner vessels (kraters, jugs, 50 Yon 2001, 123. 51 Gittlen 1981, 51. 52 Mazar 1992, 293, footnote 26; Prag 1985, 157, 159, 162. 53 Gittlen 1977 (1984), 523. 54 Hankey 1993; Leonard 1981.

Fig. 5. Mycenaean pottery 15

16 cups/stemmed cups and bowls); ritual vessels (e.g. kernos and rhyta). The figurines belong to an additional category. 55 At Tell Abu Hawam and at ancient Ugarit the open forms exceed the closed. 56 One might have expected a majority of closed forms (transport containers) at these main emporia; however the easy access to imports probably made the open dinner vessels more integrated in the various contexts of these cities. 57 Quantitative division between open and closed forms differs further from site to site as exemplified by Tel el-ajjull, where almost no open forms are found. 58 The Mycenaean pottery was valued both for its own sake and also imported as containers for other commodities like oil (e.g. in narrow-mouthed jars, flasks and stirrup jars) and unguents (in wide-mouthed jars like pyxis and alabastron). The main contents of the vessels, according to Leonard were olive oil and olive oil-based products combined with herbs and spices made by Greek specialists. Leonard also asserts that the trade in fine tableware filled gaps in the production of local pottery. 59 This is however rejected by Wijngaarden who claims that the quantities of Mycenaean pottery were too small (less than 1% even at Ugarit) to fill any gaps in the local repertoire. 60 Influx and distribution of Mycenaean pottery Prior to the Late Bronze Age, contacts between the Aegean world and the southeast Mediterranean were limited to a small-scale trade at an elite administrative level. The Minoan palace centers, Egypt and the large flourishing towns in the Levant (Ugarit, Alalakh and Byblos) dominated the interaction. 61 This state changed in the Late Bronze Age when Mycenaean pottery was spread all over the Levant and Egypt. The main influx was of LH IIIA2 and IIIB1 pottery. From this period more than one hundred sites are known in the Levant. 62 55 Wijngaarden van 2002, 14f, fig. 2.2. 56 Balensi 1980, 485; Wijngaarden van 2002, 13, footnote 32. 57 See e.g. Wijngaarden van 2002, 73. 58 Steel 2002. 59 Leonard 1981. 60 Wijngaarden van 1999, 31f. 61 Warren 1995, 2-5. 62 Hankey 1993; 1995; Leonard 1994; Wijngaarden van 2002, 31.

17 The distribution of the Mycenaean pottery shows that it was not only the coast cities that were involved in the commerce; the commodities also reached inland urban centers through a network of trade routes. 63 Wijngaarden s distribution map of 111 sites of Mycenaean finds in the Levant shows that Hazor belongs to one of the eleven sites with the most rich finds. To include the finds from the renewed excavations would make the city belong to the group of eight largest importers, together with Minet el Beida, Sarafend, Megiddo, Amman and Lachish, after the cities Abu Hawam and Ugarit. 64 Thus, four out of eight are actually located in the interior. This suggests that Abu Hawam and Ugarit functioned as main importers both of Mycenaean and Cypriot pottery, and were primary nodes of the exchange with the hinterland, based on the large quantity of pottery (more than 500 Mycenaean finds) and the wide range of forms found there. 65 However, they were not the only entries to the Levant, and furthermore their importance was not consistent throughout the whole LB period. 2.2.3 Provenance Studies and Imitations To establish the provenance of the pottery and to separate the originals from the locallymade imitations, a special method can be used called Neutron Activation Analysis (NAA), which matches the composition of the pottery with the clay of its origin. The method was used, for example, to establish the origin of a chariot crater and other Mycenaean vessels in tomb 387 at Laish/Dan. The result showed that this pottery derived from one region in Greece, Mycenaea/Berbati, 66 which supports Biran s suggestion of a Mycenaean colony in Dan, especially considering the context of the vessels a two generation tomb (beehive-shaped, known from Greek Argolid) with 40 skeletons and Mycenaean grave goods. 67 Although it might be hard to prove a Mycenaean trade diaspora, it seems plausible that the inhabitants of this grave had some kind of connection to trade. Another analysis shows that some Mycenaean pottery from the so- called Persian Garden tombs in the vicinity of Tel Akko originated in Nichorie on Cyprus. 63 Wijngaarden van 2002, map 6. 64 Wijngaarden van 2002, map 7. 65 Balensi 2004; Balensi, et al. 1993; Yon, et al. 2000. 66 Gunneweg, et al. 1992; Gunneweg & Michel 1999. 67 Biran 1970.

18 Additional imports, stone weights and luxury items among the grave goods, suggest that some of the deceased might have been related to merchantry. 68 However, the most important contribution of this method is the appearance of different origins of Mycenaean pottery in the Levant, which appears to have been made in workshops at Argolid, Cyprus and the Levant. This phenomenon is suggested to have chronological ramifications and/or depend on trade alliances (see below), and will of course have further implications for the interpretation of interaction studies in the Eastern Mediterranean. The widespread imitation of Cypriot pottery in the Levant is a well known phenomenon. K. Prag claims that imitations follow the dates of importation and production. In other words, imitation was not an activity due to lack of imported goods, as earlier stated. 69 She also reflects that White Slip wares appears to have been imported for aesthetic reasons and is also the least imitated. In imitation the aesthetic considerations were not the major factor and there seems to have been a strong tendency to adapt the foreign forms to the local tradition. The potters adopted/adapted for instance only those features which they could easily produce within their own craft tradition. They never produced handmade vessels, they did not push the handle through the body of the pot, they mostly avoided the round-based shape and continued to prefer red paint and local panel styles of decoration. In other words, this reflects a kind of hybridization. According to Prag, Cypriot pottery found in cultic context is almost exclusively imports. 70 However, regarding the Mycenaean ware an increase in imitations coincides with the fall-off of this pottery in LB IIB / Iron I Levant. This decrease depended, according to Hankey, on a scarcity of supplies in the Aegean which also affected the Mycenaean pottery imports to Cyprus. The imitations of these imports, which had become a part of daily life, were meant to solve this problem. 71 A. Killebrew instead explains the end of importation and concurrent increase in imitation as the last of three phases in the organization of the trade: The first phase 68 Ben Arieh & Edelstein 1977, 19 footnote 46; Gunneweg & Michel 1999, 995. 69 Prag 1985, 162f. 70 Prag 1985, 160-163. 71 Hankey 1993, 103.

19 reflects the concentration of pottery originating in the palatial centers in the Argolid region and consisted of LH IIA2-IIIB containers of precious oils and ointment and a smaller number of fine table ware. The second phase, represented by LH IIIB pottery, which is the end of LB when the Greece palatial centers declined and new multiple manufacturing centers emerged, points to a taking over by peripheral groups. This phase included fewer and more declined forms (stirrup jars and flasks) from multiple workshops (acc. to NAA), representing the diffusion of the west Aegean style to the east and hence a trade, although less intensive, including eastern Aegean, the southern coast of Turkey and Cyprus. A third phase is Mycenaean IIIC:1b at the end of LB IIB/Iron I, which appeared first on the coastal cities in the Levant and replaced the indigenous pottery. Their shape and technology are Aegean, with some Cypriot inspired forms, but they are locally produced and reflect more everyday use. This last phase indicates, according to Killebrew, the incursion of new groups of people with probable origin in Cyprus, Rhodes and/or southern Anatolia. 72 The Origin of the Bichrome Ware The origin of the Bichrome Ware, which occurs in late MB and early LB in the Levant, 73 has rendered some problems among scholars. 74 C. Epstein made the first serious study of this pottery type in her PhD. She regarded it as a mainland ware influenced by the Hurrians. 75 NAA studies, conducted by M. Artzy et al, instead point to eastern Cyprus, particularly Milia, as the provenance of this pottery. 76 However the provenance of Bichrome Ware is more complicated than that. Recent excavations on Cyprus have shown that this pottery is not only concentrated on the east side but spread all over the island, and maybe especially on the western part. 77 Further, it is suggested that this ware which consists of two categories, one very fine ware with neatly motif-painted decorations covering the surface and a second inferior made of gritty clay simply 72 Killebrew 1998, 160-166. 73 Prag 1985, 161; Åström 2001a, 136. 74 See e.g. Åström 2001b. 75 Epstein 1966. 76 Artzy 2001; Artzy, et al. 1973; 1975; Artzy, et al. 1978. 77 Karageorghis 2001, 144.

20 decorated on part of the body, should be classified as two separate pottery types. 78 There are different opinions on the original source of inspiration of the motifs of the fine ware; Anatolia, Syro-Mesopotamia, the Aegean and Egypt are suggested. 79 What seems to be concluded though is that the original main production of the Bichrome Ware came from Cyprus, but local imitations are prevalent in Syria-Palestine and Egypt as well. 80 The distribution of Bichrome Ware in Palestine shows that it is concentrated in the coastal cities and along important inland routes such as the main north south trade route passing Hazor, as well as the Jezreel Valley route from the sea to the interior. 81 3 THE DISTRIBUTION OF IMPORTED POTTERY AT HAZOR 3.1 Area Descriptions The areas of the Upper City of Hazor consist of Area A, B, BA, G, L and M (Fig. 2), of which only Area A and M contain substantial LB structures. The Northern Temple of Area A, its courtyard and surroundings, as well as the NE corner of the Southern Temple were excavated in the 1960 s. The finds and plans of the former temple are published and fully included in the investigation below. The Southern Temple though was completely unearthed in the renewed excavations which started in 1990 and are still in progress. These excavations focus on the tel and have yielded several monumental structures in Area A (A1-A6). A large Ceremonial Palace with a courtyard constitutes the focal point on the summit of this area. This structure should rather be seen as a royal/religious building where cultic and ceremonial activities were performed, rather than as an administrative palace. 82 Such an administrative palace is instead indicated by a podium complex found in Area M on the northern slope of the tel. This complex most likely constituted the entrance of a huge Late Bronze Age administrative/living palace, which is still hidden in the dirt of the tel. 83 The pottery of these recently excavated areas, in Area A and M, is not yet published and will therefore not be included in the tables and 78 Karageorghis 2001, 148f. 79 Artzy 2001, 163; Karageorghis 2001, 152f. 80 Artzy 2001, 168; Artzy, et al. 1978; Bietak 2001, 175; Hein 2001, 140f. 81 Artzy 2001, 167. 82 Ben-Tor 2006; Zuckerman 2006. 83 Zuckerman 2003b; 2006.

21 diagrams below. However, these areas are to a certain extent considered in the discussions. The areas of the Lower City are Area C, D, E, F, H, K and P (Fig. 2). The two latter, K and P, constitute city-gates with connected buildings. Area D and E are hard to interpret and only contain fragments of badly preserved walls. A couple of kilns in Area D hint at some kind of industrial activities. Several large cisterns, originally used as water reservoirs but later reused as silos and for burials, are most noticeable in these two areas which have yielded a remarkably large amount of imported pottery. Area C included dwelling quarters, a temple and a couple of pottery workshops. Area F also included a dwelling quarter and a probable cult place; this area yielded the most imported pottery of all the areas so far reported at Hazor. The implication of this phenomenon will be further discussed below. Area H consisted of succeeding (from the end of MB until the end of LB) monumental temples with courtyards and adjacent buildings. 84 All the imported pottery found in this area is, in this investigation, regarded as being connected to the temple in the respective period. Table I and II (see appendix), show the Cypriot and Mycenaean pottery found in the different areas at Hazor during LB. The distribution of these types in the areas of the Upper and respective areas of the Lower City are displayed in the diagrams below, divided into their find contexts in LB I and LB II (Fig. 6-12). As seen, no Mycenaean pottery is found in LB I contexts, which concurs with the general picture of the importation of this pottery type into Syria-Palestine. 85 Bichrome pottery does, as expected, only occur in LB I, except for two sherds in Area A, which appeared in disturbed LB II layers. The difficulties in several areas of establishing the division between level 1A (LB IIB) and 1B (LB IIA) prevent a more precise division of the pottery within LB II. However, about 75% of the LB II Cypriot as well as 70% of the LB II Mycenaean pottery should be ascribed to LB IIA. 86 This statement though has to be updated when the imported pottery of the renewed excavation has been published, of which most originated from the destruction level (LB IIB) of Area A and M and include 84 Yadin, et al. 1989, 215-264. 85 Wijngaarden van 2002. se också Leonard, Killebrew etc 86 According to an estimation made by the author.

22 approximately 150 Mycenaean and 220 Cypriot sherds of hard-identifiable and fragmented character (see below Chapter 3.3.1). 87 3.2 THE LB I CYPRIOT POTTERY The total number of LB I Cypriot vessels/sherds (excluding Bichrome) is 31 pieces. As seen in Fig. 6 and Tables II (in the appendix) the most abundant Cypriot pottery in LB I appeared in Area E and H were it was found in specific contexts. CYPRIOT POTTERY LB I 12 Number 10 8 6 4 2 Temple/Cult Burial Habitation Silo (cistern) Other/indef. Area total 0 BA C E F H K L Area Fig. 6. Distribution of Cypriot pottery at Hazor during LB I Numbers 12 10 8 6 4 2 0 CYPRIOT FORMS IN VARIOUS CONTEXTS, LB I Temple/Cultic Burial Habitation Silo (cistern) Other/Indef. Contexts Bowl Monochr Bowl WS I Bowl WS II Bowl other Jug BR I Jug BR II Juglet Wh. Shaved Juglet other Context total Fig. 7. Cypriot pottery forms in different contexts at Hazor in LB I 87 Zuckerman 2003b, 269-281 (in Hebrew).

23 In Area E the LB I pottery, which consisted of bowls (Monochrome, WS I and II milk bowls ) and jugs and juglets (BR I, White Shaved), was only found in cistern L. 7021, as it seems, thrown or fallen down. 88 This pottery may have been connected to some kind of cultic activities, based on the composition of this pottery assemblage and the general character of this area, e.g. several carved cup marks on the bedrock surface. Except for a few additional jugs/juglets in Area E the Cypriot pottery types of this area are similar to the assemblage of Area H. In this latter area the LB I pottery was associated with the Stratum 2 temple, where it appeared both inside and outside this structure. 89 A somewhat smaller amount of LB I Cypriot pottery appeared in Area F where a juglet was found in a burial and two BR I jugs so called bilbils were uncovered in unstratified contexts; only one bowl (Monochrome) appeared, also unstratified. Two Monochrome bowls appeared in habitation contexts in Area C and two sherds, probably bowls, in a structure connected to the Area K gate (other/indef. context). The most common forms (Fig. 7) in LB I were bowls (Monochrome and WS I and II) followed by the bilbil jugs (BR I and II). Strangely no Cypriot pottery from this period is found in Area D. 3.2.1 The Bichrome Pottery at Hazor Neither NAA tests nor visual attempts have been made to distinguish possible local from imported Bichrome ware at Hazor. Thus, since its Cypriot origin cannot be concluded, it is treated in a separate diagram (Fig. 8). Most of the 20 sherds of this pottery type found on the site appeared in temples (A north and H) and consisted of various shapes, bowls, kraters and jugs. Three jugs and three kraters were found in a burial in Area F and the rest of the Bichrome pottery in this area appeared in unstratified levels. In Area E they were all found in cistern L. 7021 except for one that was found in an indefinable context. As mentioned above, the character of the pottery in L. 7021 hints to a special use, perhaps connected to cult. This pottery type seems to have been concentrated in cultic contexts at Hazor, a picture that differs from the Bichrome at Tell el Dab a and Ezbet Helmi where most of the Bichrome pottery is found in settlement contexts, especially in a palace 88 Yadin, et al. 1958, 148. Cistern L. 7021 is labelled as other in the diagram since its use is unknown. 89 According to an investigation in my forthcoming PhD.

24 area. 90 However, almost no habitation contexts of LB I have been excavated at Hazor, which might explain the absence of Bichrome in this situation. Thus the Bichrome pottery at Hazor including the Cypriot LB I pottery discussed above, appeared mainly in cultic circumstances. This is especially true if the Area E pottery should be attributed to this context. BICHROME POTTERY LB I Number 10 8 6 4 2 Temple/Cult Burial Silo (cistern) Other/indef. Area total 0 A E F H Area Fig. 8. Distribution of Bichrome pottery at Hazor during LB I 3.3 THE LB II CYPRIOT AND MYCENAEAN POTTERY In LB II the number of Cypriot vessels increases on the site, 107 pieces derive from this period (Table II), of which the main part, approximately 75%, is found in LB IIA strata. Continuing occurrences of this pottery can be found in the temples in Area H, and now also in Area A (the Northern Temple), where the pottery had an obvious cultic function and consisted mainly of bowls of the Monochrome and White Slip I and II types as well as some BR I and II jugs (Fig. 9 and 10 and Table II). It was used also in smaller cultic connections in the Area K gate and in the open cult place of Area F. In Area C, on the other hand, no Cypriot pottery was found in connection to the temple, but instead a few WS bowls were found in habitation contexts. However, about a third of all Cypriot pottery in this period derives from Area F, of which most appeared in burials, L. 8144-8145 and L. 8065 and consisted of mostly jugs of the BR I and especially of the BR II type. 90 Bietak 2001, 175; Hein 2001, 237-241.

25 As seen in the diagram (Fig. 10), the most common form is the WS bowl in cultic contexts, followed by BR jugs in burials. The noticeably high amount of Cypriot pottery found in other/indefinable contexts derives mainly from unstratified levels and unidentified contexts in Areas A, C, D and F. CYPRIOT POTTERY LB II Number 40 35 30 25 20 15 10 5 0 A BA C D E F G H K Area Temple/Cult Burial Habitation Silo (cistern) Workshop Other Total Fig. 9. Distribution of Cypriot pottery at Hazor during LB II CYPRIOT FORMS IN VARIOUS CONTEXTS LB II 40 35 30 Bowl Monochr Number 25 20 15 10 5 0 Bowl WS I/WS II Bowl BR I/II Tankard BR II Jug/Juglet BR I/II Other juglets Temple/Cultic Burial Context Habitation Silo (cistern) Workshop Other/indef. Total Fig. 10. Cypriot pottery forms in various contexts at Hazor in LB II

26 The strata of LB II have yielded 59 sherds and vessels of Mycenaean pottery. Some interesting aspects can be observed regarding the distribution of Mycenaean pottery on the site (Fig. 11, 12 and Table I). Area F contained more than twice as much of this MYCENAEAN POTTERY LB II Number 30 25 20 15 10 5 0 A B BA C D F H P Area Temple/Cult Burial Habitation Silo (cistern) Workshop Other TOTAL Fig. 11. Distribution of Mycenaean pottery at Hazor in LB II MYCENAEAN FORMS IN VARIOUS CONTEXTS 20 Flask Number 15 10 5 0 Stirrup jar Pyxis/Alabastron Piriform jar Krater Cup/Kylix Temple/Cult Burial Habitation Silo (cistern) Workshop Other/indef. Bowl Figurine Unidentified Context Total Fig. 12. Mycenaean pottery forms in various contexts in LB II pottery type than the next areas, which are Area A and D, followed by Area C. Most of the Mycenaean pottery in Area F derives from the same burial, L. 8144-8145, as the Cypriot imports. This tomb of multiple burials, located just outside the dwelling quarter, is conspicuously rich in finds. Unlike the WS bowls of Cypriot origin, the Mycenaean

27 pottery in this burial consists mainly of containers such as flasks, stirrup jars and pyxides. This together with the remarkable abundance of Mycenaean pottery in and around the rather large and well built houses in this area hints to its inhabitants wealth and easy access to imports. 91 Except for an unidentifiable sherd in house L. 6022 in Area C, no other area has yielded any Mycenaean pottery in habitation contexts. Area D also included a burial with Mycenaean pottery. This was located in a cistern, L. 9027, which yielded three pyxides, an amphoroid krater and a deep bowl. Another cistern, L. 9017 interpreted as a silo, included a couple of Mycenaean sherds. Van Wijngaarden observes the atypical nature of the assemblage in this silo and suggests its connection to a nearby kiln and perhaps some industrial activities. 92 A few sherds were also found in unstratified levels in this badly preserved area. As with the Cypriot pottery the Mycenaean examples in Area A and H are connected to the temple. Interesting is Area C where most of the Mycenaean sherds were found in industrial contexts in the workshops. These vessels included pyxides and a piriform jar, which might have been connected to the pottery production. 93 3.3.1 Some Notes on the Imported Pottery of the Renewed Excavations The pottery is derived from the destruction level (XIII and 1A), LB IIB of Areas A and M. This imported pottery is not yet published, but briefly discussed in an unpublished Ph.D.,dissertation, hence it is impossible to include it in the diagrams. 94 Area A The Ceremonial Palace About 90 Mycenaean sherds appeared in the areas of the Ceremonial Palace and its courtyard, many of which were found in recent contexts of the Iron Age. The sherds were in most cases very small and hard to identify and no complete vessels were found. The shapes are typical of the LH IIIA2 ware and continue to appear in the LH IIIB1. The quality of these vessels is high and the pottery in general is similar to the Mycenaean 91 For a description of these houses and those in Area C see Daviau 1993, 219-252; Foucault-Forest 1996, 67-73; Yadin, et al. 1960, Chapter 4 and 5. 92 Wijngaarden van 2002, 91. 93 Wijngaarden van 2002, 89f. 94 The description below of the pottery in the Area A and M of the renewed excavations is based on Zuckerman 2003b, 269-281.